JP2007526699A - Method for reducing waiting time in push-to-talk service - Google Patents

Abstract

A method of processing user speech data for transmission to a participant or participants in a Push to talk Over Cellular (PoC) session over a cellular telephone network. The method comprises, detecting an initial period of silence in the initial talk burst of the session, and removing that period of silence from the speech data prior to replaying of the speech data to the or each other participant. These signal processing steps may be carried out at one of the initiating terminal, the receiving terminal, or the IMS core.

Description

  The present invention relates to a method for reducing latency in a push-to-talk service. In particular, the present invention relates to a method for reducing latency in a so-called push-to-talk service via a cell.

  Push-to-talk is the general name for a service that allows mobile radio handset users to communicate with each other almost instantaneously by pressing one button or at least a few buttons. The industry is in the process of standardizing push-to-talk services for introduction into current and next generation cellular networks, including GSM with packet data services and 3G. The service is known as “Push-to-Talk over cells” (PoC).

  PoC uses the IP Multimedia Subsystem (IMS) standardized by the 3rd Generation Partnership Project to facilitate the introduction of advanced data services, especially real-time multimedia services, into cellular networks. IMS follows the Session Initiation Protocol (SIP) defined by the Internet Engineering Task Force (IETF) for setting up and controlling multimedia IP-based sessions. FIG. 1 is a diagram schematically showing a configuration of a cellular network that provides a PoC service among several user terminals or user equipment (UE) 1 in 3G. The UE 1 is connected to each radio access network 2. The radio access network 2 is similarly connected to the IMS core 3. There are several servers in the IMS core 3. The server includes a visited call session control function (S-CSCF) server 4 that is a main SIP server that maintains the session state of the IMS service. In addition, the IMS core 3 includes a proxy call session control function (P-CSCF) server 5 that is a first contact point for the UE 1 and transfers a SIP message to the S-CSCF. Servers in the IMS core 3 are distributed within the network of the operating company and among a plurality of networks. Furthermore, the PoC server 6 is located in the IMS or connected to the IMS. The PoC server 6 may incorporate a media resource function (MRF) as defined by 3GPP.

  FIG. 2 is a diagram illustrating certain signal control associated with setting up a PoC session over the network of FIG. 1 (not shown, but additional messages may be transmitted between various nodes). ). The subscriber starts a session by pressing the appropriate button on his terminal UE # 1. As a result, a SIP request (SIP_INVITE) message is transmitted to the other terminal UE # 2 via the PoC server of the IMS core. Thereafter, signal transmission is further continued between the terminal and the IMS. As described above, the main problem of PoC is the connection between terminals that is almost instantaneous. Thus, large delays in the transmitted voice should be avoided.

  The time between when the SIP request message is sent and the IMS receives the response from the callee is about 3 seconds due to the intrinsic characteristics of network establishment (eg, paging, establishment of temporary block flow (TBF), etc.) It can be. In order to speed up the initial connection process, the originating (calling) subscriber can initiate a call when the terminal receives a SIP 202 accept message from IMS, even though the called party has not yet accepted the session. The receipt of the acceptance message is usually communicated to the initiating subscriber by playing a sound or “beep” at the terminal. The initial voice burst may be buffered by the PoC server until the PoC server in the network receives the SIP 200_OK message from the partner (calling) terminal. When the message is received, the voice burst is immediately transmitted to the partner terminal. Nevertheless, the delay felt by the called party remains significant and it is desired to further reduce the delay.

  The inventor of the present invention recognizes that the initiating subscriber is unlikely to initiate a call even immediately after the sound is played, due to both the subscriber's reaction time and its “think time”. . As illustrated in FIG. 2, this delay is calculated as 0.8 seconds.

  A first aspect of the present invention is a method of processing user voice data for transmission to one or more participants in a push-to-talk session over a communication network. The method includes removing an initial silence period from the audio data prior to playing the audio data to the participant or each other participant.

  The present invention is particularly suitable for removing the initial silence period from the initial speech burst provided by the initiator of a push-to-talk session. This has the effect of reducing the delay between the generation of the audio burst by the initiating subscriber and the playback of the audio burst to the participant or each other participant.

  The communication network is a cellular telephony network, and the push-to-talk service is preferably a push-to-talk (PoC) service via a cell.

  Embodiments of the invention may include analyzing audio data and identifying an initial silence period. This step may be performed at the initiating terminal, the node in the communication network, or the receiving terminal. Similarly, the step of removing the silence period detected from the transmitted voice data may be performed at the initiating terminal, the node in the communication network, or the receiving terminal. If the communication network is a cellular telephone network and the push-to-talk service is a PoC service, the network node is preferably present in the IP Multimedia Subsystem (IMS).

  If the detection and removal steps are performed at the initiating terminal, the detecting step may include analyzing audio data during or after data recording at the initiating terminal.

  Embodiments of the present invention may include monitoring the sound level and starting recording audio only when the sound level exceeds a predetermined threshold. This step may be performed at the initiating terminal or a server node in the communication network. In another embodiment of the present invention, an initial period that is believed to include silence is predetermined, and the beginning portion of the audio data is clipped to remove the predetermined period. The predetermined period may be fixed or may be adapted based on the user's call / use pattern.

  The step of removing the initial silence period from the audio data may be performed in real time as the audio data is received, or may be performed by later processing the stored or buffered audio data.

  A second aspect of the present invention is a server node used in a communication network that provides a push-to-talk service to subscribers. The node includes a receiver that receives a burst of audio data from a participant in a push-to-talk session. The node also includes a processor that detects an initial silence period of the audio data burst and removes the silence period detected from the audio data prior to transmission to a session participant or each other participant.

  The server node is preferably located in an IP multimedia subsystem in a cellular telephone communication network. Further, the server node preferably includes an interface to one or more Session Initiation Protocol (SIP) servers having a Visited Call Session Control Function (S-CSCF) server.

  A third aspect of the present invention is a mobile terminal used in a communication network that provides a push-to-talk service to a subscriber. The mobile terminal includes a receiver that receives audio data from a user of the terminal and a processor that removes a silence period from the audio data prior to transmission to a session participant or each other participant.

  The mobile terminal is preferably a wireless terminal, and the communication network is preferably a cellular telephone network that provides a push-to-talk service over a cell.

  The mobile terminal may be a terminal used by a user of the terminal or another terminal participating in the session.

  Delays specific to establishing a push-to-talk (PoC) session over a cell have been described above with reference to FIGS. In the following, a mechanism for sufficiently reducing these delays will be described with reference to some possible embodiments. This embodiment is based on the fact that it does not start a call until some time has passed after the terminal indicates that the PoC session participant may start the call by sounding or by other means. Yes.

  In the first embodiment of the present invention, the media resource function (MRF) of the PoC server starts receiving the initial voice burst transmitted from the initiating subscriber's mobile terminal (UE # 1) following the start of the PoC session. To do. This burst includes an initial silence period, which may last for example 0.8 seconds, or background noise, and is transmitted in several real-time protocol (RTP) frames from UE # 1 to the PoC server. The PoC server buffers the received voice data and waits for reception of a SIP 200_OK message (acceptance from other participants) from other participants in the session. This can take from a few milliseconds to a few seconds. During this time, the PoC server analyzes the buffered data to determine the length of the initial silence period and cuts the data to remove that period once identified. Following the reception of the SIP 200_OK message, the PoC server starts transmitting the voice cut from the head of the buffer.

  FIG. 3 is a diagram illustrating signal control related to a session procedure. As described above, the PoC server of the IMS core calls the called party (here, a case where there are only two participants is illustrated). On the other hand, the PoC server simultaneously gives the “speaking right” to UE # 1 (MS # 1). By removing the initial silence period from the voice burst, the voice is received by UE # 2 (MS # 2) 0.8 seconds earlier than otherwise. The entire session is advanced by the same amount of time. Therefore, it will be understood that the real feeling of the participants is enhanced.

  The process of determining the existence and duration of the initial silence period is performed by analyzing the volume of the audio signal received at the PoC server. When the volume exceeds a predetermined threshold, it is assumed that the call has started and the silence period has ended. Of course, more sophisticated algorithms may be used. For example, the audio signal may be analyzed by the presence of a call specific pattern. As a result, the presence of background noise is suppressed because an incorrect call instruction is given. Another approach assumes that the call cannot begin for a fixed period after the sound is heard, for example 0.8 seconds, and removes that period from the beginning of the speech burst. Depending on the initiator (calling party) behavior or the statistically analyzed behavior of the subscriber group, the length of this period may be dynamically adapted.

  The approach described above is based on speech analysis procedures performed within the IMS core and silence period elimination. Providing sufficient processing power to achieve this approach is not a problem. However, these steps may be performed at the initiating terminal as long as sufficient processing power is available at the initiating terminal. That is, immediately after appropriate sounding at the initiating terminal, the initiating terminal analyzes the user's voice and determines the length of the initial silence period. In addition, at the initiating terminal, a sound may be played prior to receiving the “call instruction” message from the IMS core.

  Analysis and modification of the initial speech burst may instead be performed at the receiving terminal (or multiple receiving terminals if the session includes more than two participants). However, it is necessary that the data transmission rate via the interface between the receiving terminal and the IMS core is sufficiently faster than the call rate, and that the “development” of the received voice is in time before playback. In such a case, the session delay will be sufficiently reduced by the detection and removal of the initial silence period, even if not as effective as achieved by the other solutions described above.

It is a figure which shows the outline | summary of the communication network in the cellular telephone which provides the subscriber with the push-to-talk service via a cell. FIG. 7 illustrates signal control associated with a setup phase of a push-to-talk session over a cell and an initial voice burst. FIG. 6 illustrates the improved setup phase of a push-to-talk session over a cell and signal control associated with an initial voice burst.

Claims (18)

A method for processing user voice data for transmission to one or more participants in a push-to-talk session over a communication network comprising:
Removing the initial silence period from the audio data prior to playing the audio data to the participant or each of the other participants. The audio data is
The method of claim 1, wherein the initial speech burst provided by the push-to-talk initiator. The communication network is a cellular telephone network;
The method according to claim 1 or 2, wherein the push-to-talk is a push-to-talk service via a cell. 4. A method according to any of claims 1 to 3, further comprising analyzing the audio data to identify the initial silence period. Analyzing and removing the audio data to identify the initial silence period;
5. A method according to claim 2 when dependent on claims 2 and 3, characterized in that an acceptance from the participant or each other participant is performed prior to being accepted in the communication network of a session. . Analyzing the audio data to identify the initial silence period;
The method according to claim 4 or 5, wherein the method is executed at one of the initiating terminal, a node in the communication network, or a receiving terminal. Removing the initial silence period from the transmitted audio data;
7. The method according to claim 1, wherein the method is executed by the initiating terminal, a node in the communication network, or a receiving terminal. The method according to claim 7, when dependent on claim 3, wherein the node in the network is a node having a media resource function. 8. A method according to claim 7, when dependent on claim 3, wherein the nodes in the network are located in an IP multimedia subsystem (IMS). 10. A method as claimed in any preceding claim, further comprising the step of monitoring the sound level to determine that a call has been initiated. Pre-defining an initial period in anticipation of including silence;
10. The method according to claim 1, further comprising a step of cutting off a beginning portion of the audio data excluding the predetermined initial period. 12. The method of claim 11, wherein the predetermined initial period is fixed or adapted according to subscriber behavior. A server node for a user in a communication network providing a push-to-talk service to a subscriber,
A receiver for receiving a burst of audio data from a participant in a push-to-talk session;
Detecting an initial silence period in the audio data burst and prior to transmitting the audio data burst to a participant or each other participant in the push-to-talk session, the detected initial silence from the audio data burst. A server node comprising a processor for removing a period. Located in an IP multimedia system in a cellular telephone communication network,
14. The server node of claim 13, further comprising an interface to one or more session initiation protocols (SIP) having a visited call session control function (S-CSCF) server. A mobile terminal for a user in a communication network that provides a push-to-talk service to a subscriber,
A receiver for receiving voice data from a user of the terminal;
A mobile terminal comprising: a processor for removing a silence period from audio data prior to transmitting the audio data to a participant in a session or each other participant. The mobile terminal is a wireless terminal;
The mobile terminal according to claim 15, wherein the communication network is a cellular telephone network that provides a push-to-talk service via a cell. The receiver
17. A mobile terminal according to claim 15 or 16, comprising means for converting speech into an analog or digital electrical signal. The receiver
17. The mobile terminal according to claim 15 or 16, further comprising means for receiving voice data generated by a counterpart mobile terminal via an interface such as the communication network.

Images